Lecture 6

Question 1

Pretest-Posttest with control

Answer 1

2 or more groups formed by random assignment, with differing levels of I.V, RCT

Question 2

Posttest only design

Answer 2

2 or more groups formed by random assignment, with differing levels of I.V, time related variables (generally)

Question 3

Factorial Design

Answer 3

2 or more I.V with subjects randomly assigned to various combinations of levels

Question 4

Randomized Block Design

Answer 4

Identical to factorial except the blocking factor is a confound, rather than an experimental factor of theoretical insight

Question 5

One-way repeated measures design

Answer 5

Repeated measures involve exposing participants to all levels of I.V, may present an order effect

Question 6

Crossover Design

Answer 6

A repeated measures design that counterbalances measures to control order effect

Question 7

Quasi-experimental design

Answer 7

An experimental procedure is applied but not all extraneous variables are controlled

Question 8

One group pretest-protest design

Answer 8

1st set of measurements taken before & after treatment on one group of subjects (no control)

Question 9

Non-equivalent groups

Answer 9

Most common quasi-experimental, groups are formed by something other than random assignment

Question 10

Internal Validity

Answer 10

Is relationship causal? Temporal precedence (cause occurred before effect). Central issue is control of confounds (alternative explanations)

Question 11

Covariation of C&E

Answer 11

If X, then Y, If not X , then Not Y –> Establish relationship between IV & RV

Question 12

1 Threat of Indiv Differences

Answer 12

Extraneous variation, in the form of indiv differences, can severely bias experiments

Question 13

2 Threat of indiv differences

Answer 13

Homogenous subject selection –> Eliminate confounding variables by not allowing subjects to vary on the variable of interest

Question 14

3 Threat of indiv differences

Answer 14

Matching –> Groups can be made eq on confounding variables by ensuring that groups are matched before experiment begins

Question 15

4 Threat of indiv differences

Answer 15

Repeated Measures –> Involve using subjects as their own control

Question 16

5th Threat of indiv differences

Answer 16

Blocking / Analysis of Covariance –> Evaluate a confound as an additional factor in the experiment –> Blocking = categorical variables, A of C = continuous variable

Question 17

Threats to Internal Validity

Answer 17

History, Maturation, Attrition or Mortality, Testing, Instrumentation, Regression, Selection, Diffusion of treatments, Compensatory equalization of treatment, selection history, selection-instruments

Question 18

Intention to Treat

Answer 18

Occurs when a participant drops out of the study or chooses to be in a different treatment group –> Intention to treat = treat data as part of the group the participant was assigned.

Question 19

Threats to external validity

Answer 19

Selection (sample specific), Setting (context-specific_), History (results due to unique sample experiences)

Question 20

Birth Rate

Answer 20

of live births during the year / total population at mid year

Question 21

Mortality Rate

Answer 21

of deaths during a specific time period / total pop at the midpoint of the time period

Question 22

Prevalence

Answer 22

of existing cases at point in time / total population

Question 23

Cumulative Incidence

Answer 23

of new cases over a period in time / total pop

Question 24

Rule of 3

Answer 24

When N > 30, we’re 95% certain the events in question will not occur more than 3/N of the time

Question 25

Relative Risk Ratio

Answer 25

Risk of disease among exposed / risk of disease among unexpected –> How much does risk factor increase probability of developing a disease –> Only in Cohort Studies

Question 26

Odds Ratio

Answer 26

How much more likely is it that an individual with the disease will have a particular risk factor? = Ratio of exposure odds among cases to exposure odds among controls –> Valid for case-control studies